RHIZOSOLENIA CURVATA 441 



these are replaced to an extent varying with the time of year by more southerly forms, of 

 which the chief are: 



Fragilariopsis antarctica (Castracane) Hendey Aster omphalus challengerensis Castracane 

 Nitzschia seriata Cleve 1 Thalassiothrix antarctica (Schimper ex Karsten) 



Asteromphalus regularis Karsten 



Where Rhizosolenia curvata has been swept south of the convergence by local reversal 

 of the usual surface drift it usually finds itself in a community of these forms, together 

 with both the common southern phases of Corethron criophilum. Sometimes where 

 the mixing has been strong Rhizosolenia polydactyla may be quite numerous south of 

 the convergence, but it is not so good an indicator as R. curvata owing to its much more 

 irregular distribution in time and space. 



In autumn, in the convergence region, the solenoid plankton is less well defined and 

 Ceratium pentagonum Gourret becomes important. Sometimes when the phytoplankton 

 is very sparse Rhizosolenia curvata becomes numerically dominant in late autumn. Near 

 the northern limit of its range it is sometimes found along with Ceratium fusus Ehrenberg 

 and other less numerous forms whose proper habitat is the warmer part of the sub- 

 Antarctic Zone, such as various phases of Ceratium tripos O. F. Muller. 



CONCLUSIONS 



On the basis of the evidence described in this paper it is concluded that Rhizosolenia 

 curvata is a typically sub-Antarctic oceanic plankton diatom, mainly confined to the 

 colder portion of the sub-Antarctic Zone. It is frequently carried across the Antarctic 

 convergence when mixing of the surface waters occurs, and may survive for a short time 

 in the mixed water, reproducing by vegetative division only. The rarity of large in- 

 dividuals, such as would result from recent auxospore formation, on the Antarctic side 

 of the convergence, suggests that it could never persist there, for the sub-Antarctic 

 element in the mixed water becomes cooled and dissipated rapidly, while the biological 

 environment is also very different. It is thought that the occurrence of this species on 

 the Antarctic side of the convergence is a good indication that mixing is in progress or 

 has very recently taken place. The usual causes of such mixing are local wind currents 

 in the warmer months of the year, and local reversal of the usual north-easterly surface 

 drift, due to the sinking of intensely cooled surface water to the southward, in winter. 



The probable exceptions to this rule arise in the few areas (e.g. South Georgia) where 

 shoal water at no great distance south of the convergence may lead to vertical mixing, 

 and thus to reappearance at the surface of individuals that have been carried south in the 

 sub-surface current of the warm deep water. 



Fig. 7 may be taken as a summary of the distributional evidence, it includes nearly all 

 the positive records of the species, and the more significant negative observations. The 

 only records omitted are the few small hauls round South Georgia, and a few from the 



1 Nearly cosmopolitan, but reaching its maximum numbers in the Southern Ocean within the Antarctic 

 Zone. 



